Diagnostic concepts for smart device-level networks

InTech, Oct 1997 by Dierauer, Peter P

The number-one issue facing today's industrial control industry is lost productivity due to equipment downtime. Many manufacturers and OEMs believe they can reduce downtime dramatically with diagnostics. To minimize downtime, end users and suppliers must understand how diagnostics are defined and the trends developing related to smart device-level networks.

Basically, diagnostics identify something that does not work as it should, including its location and extent. This is much more information than a traditional single point-to-point wiring approach can handle. Thus, to use diagnostics, an industrial network is necessary. A network allows diagnostic information to be communicated or transferred in the right way at the right time.

What different types of diagnostics are there? There are two major points of view that are interconnected.

The first view splits diagnostics into preventive (prefailure) and reactive (postfailure) information, as shown in Figure 1. A preventive diagnostic system generates an alarm before an actual failure occurs that indicates a part of the system is getting close to a criticai condition. This gives operators an opportunity to fix a problem before the system goes down.

Reactive diagnostic information identifies existing problems.

The second view organizes the information based on its origin in a control system-process, device-level, and network diagnostic information-as shown in Figure 2. Process diagnostic information, which is application or market specific, is generated by the logical combination of sev eral input devices that identify problems on the system level (e.g., jam detection on a conveyor system).

Device-level diagnostics deliver information about the health of a device and its environment. For example, a pressure regulator cannot deliver the requested output pressure if the input pressure is below the requested level. It also cannot satisfy the system request if its activation coil is broken.

To use diagnostic information from the process and device levels, a reliable communication system is needed. Therefore, the information transport system must be working properly and any glitches must be reported. This is the third level of diagnostics at the network level. Network diagnostics include network configuration information (e.g., lost device or wrong device) as well as simple shorts or breaks of communication wires.

Most of today's industrial control applications involve reactive diagnostics implemented at the process and network level, as shown in Figure 3. However, end users are pushing hard for accurate and detailed preventive diagnostic features at a reasonable cost. Two important cost factors driving this trend include the complexity of the main control program and the number of devices in a system.

To reduce the cost and complexity of the control program and add more diagnostic information to a system at the same time, logic related to diagnostics must be distributed down to the devices. The devices. The devices themselves must gather and evaluate the necessary data to generate accurate and pre cise diagnostic information.

Another way to reduce cost is to limit the number of devices connected to the network. Today, many basic sensors are added to systems to evaluate the environmental conditions of various devices. By implementing device-level diagnostics, these basic sensors can be integrated in the same housing with the primary process control device. This reduces the overall system part costs and allows more advanced features and algorithms to be implemented to generate preventive diagnostic information.

A strong trend is developing toward smart devices with advanced devicelevel diagnostic capabilities connected to a network, which supports the concept of event-driven diagnostic information exchange, as shown in Figure 4. This trend will allow users to minimize downtime and increase productivity.